Smartphone Sensor-Based Activity Recognition by Using Machine Learning and Deep Learning Algorithms

Liu, Qingzhong; Zhou, Zhaoxian; Shakya, Sarbagya Ratna; Uduthalapally, Prathyusha; Qiao, Mengyu; Sung, Andrew H.

doi:10.18178/ijmlc.2018.8.2.674

Cited by 16 publications

(8 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At last, the model with area data shows a precision of 95% and the model without area data does 90% exactness. Smartphone Sensor-Based Activity Recognition by Using Machine Learning and Deep Learning Algorithms [6] proposed a method for activity recognition in two methods. In method1, accelerometer gives crude sensor information, which perceives the activities and the course of cell phones showing the gyrator faculties rotational developments, which are huge to recognize by individuals.…”

Section: Related Workmentioning

confidence: 99%

Untitled

2021

International Journal of Advanced Research in Computer and Comm

View full text Add to dashboard Cite

Smart phones are the most helpful apparatuses of our day-by-day life and with the propelling innovation; they get fit systematically to address user's issues and desires. To make these contraptions more useful and amazing, originators add new modules and gadgets to the equipment. Sensors has a major part in making cell phones more practical and mindful of the climate subsequently most smart phones accompany distinctive inserted sensors and this makes it conceivable to gather tremendous measures of data about the client's day by day life and exercises. The goal of Human Activity Recognition is to identify the activities performed by an individual from a given set of the information about him/her and his general environment. A great deal of exploration is being done in the field of Human Activity Recognition which human conduct is deciphered by reasoning highlights got from development, place, physiological signs and data from environments. The propose system presents a deep learning framework for human activity recognition using a smartphone data. The dataset was downloaded from kaggle.com. The dataset contains accelerometer and gyroscope data gotten from a Samsung Galaxy S2 smartphone. The accelerometer and gyroscope data is made up of different activities performed by an individual. The propose system uses keras framework and Theano as backend in build our model. After successful training, the proposed method had an accuracy of 99.06% on the 120 th epoch.

show abstract

Section: Related Workmentioning

confidence: 99%

Untitled

2021

International Journal of Advanced Research in Computer and Comm

View full text Add to dashboard Cite

show abstract

“…In [12] CNN was applied on a dataset collected from a smartphone. Firstly, they analyzed the performance of machine learning algorithms.…”

Section: Related Workmentioning

confidence: 99%

“…A comparison of related studies in terms of architectures, activities and best performance results are summarized in Table I 1 . In most of the studies, the CNN architecture is used [8,[11][12][13]. The studies that use a different dataset than ours focus on a smaller number of activities.…”

Section: Related Workmentioning

confidence: 99%

“…In contrast, in this paper, we evaluate the impact of the three popular deep learning architectures, which are CNN, RNN and LSTM. Moreover, we investigate the use of magnetometer sensor in combination with accelerometer and gyroscope sensors whereas some of the studies only focus on the use of accelerometer [9,11,13] or accelerometer and gyroscope [4,8,12].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Evaluation of Deep Learning Models for Smoking Recognition with Smartwatch and Smartphone Sensors

Akan

Ağaç

İncel

2021

Balkan Journal of Electrical and Computer Engineering

View full text Add to dashboard Cite

Smartwatches and smartphones are extensively used in human activity recognition, particularly for step counting and daily sports applications, thanks to the motion sensors integrated into these devices. Machine learning algorithms are often utilized to process sensor data and classify the activities. There are many studies that explore the use of traditional classification algorithms in activity recognition, however, recently, deep learning approaches are also receiving attention. In this paper, we utilize a dataset that particularly consists of smoking-related activities and explore the recognition performance of three deep learning architectures, namely Long-Short Term Memory (LSTM), Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN).We evaluate their performances according to different hyperparameters, different sensor types and device types. The results exhibit that the performance of LSTM is much higher than that of CNN and RNN. Moreover, the use of magnetometer and gyroscope together with accelerometer data improves the performance. Use of data from smartphone sensors also enhances the performance results and the final accuracy with the best parameter combinations is observed to be 98.2%.

show abstract

“…In order to handle streaming of data, sliding window technique is used. Qingzhong et al [5] proposes a method for activity recognition in two steps.…”

Section: Literature Reviewmentioning

confidence: 99%

Design of LSTM-RNN on a Sensor Based HAR using Android Phones

Arthishwari*,

Anand

2020

IJRTE

View full text Add to dashboard Cite

Activity Recognition (AR) is monitoring the liveliness of a person by using smart phone. Smart phones are used in a wider manner and it becomes one of the ways to identify the human’s environmental changes by using the sensors in smart mobiles. Smart phones are equipped in detecting sensors like compass sensor, gyroscope, GPS sensor and accelerometer. Human Activity Recognition (HAR) framework collects the raw data from sensors and observes the human movement using different classification methods. This paper focuses for Activity Recognition (AR) based on smart phone by analyzing the performance of various Deep Learning (DL) approach using in-built gyroscope and accelerometers. In this work, HAR dataset can be utilized from UCI based Machine Learning repository. The sensors such as gyroscope and accelerator are used to record the signals and performs various activities namely walkingdownstairs, walking-upstairs, jogging, standing, walking and sitting while a user wearing the smartphone in a pocket. The performance metrics has analyzed to recognize user’s activities using DL approach namely Recurrent Neural Network with Long-Short Term Memory (RNN- LSTM) were applied. The result provides 96% better accuracy for RNN-LSTM with minimum Mean Absolute Percentage Error (MAPE) when compare to other machine learning classifier.

show abstract

Smartphone Sensor-Based Activity Recognition by Using Machine Learning and Deep Learning Algorithms

Cited by 16 publications

References 15 publications

Untitled

Untitled

Evaluation of Deep Learning Models for Smoking Recognition with Smartwatch and Smartphone Sensors

Design of LSTM-RNN on a Sensor Based HAR using Android Phones

Contact Info

Product

Resources

About